Research Article
Medical Image Classification Based on Information Interaction Perception Mechanism
Table 3
Performance of different depth IIP-Net (%).
| Model | Accuracy | Precision | Recall | Specificity | F1-measure | Polyp Acc |
| IIP-Net54-GAP-FC | 99.59 | 99.40 | 99.40 | 99.70 | 99.40 | 99.40 | IIP-Net54-FC | 94.98 | 94.88 | 94.60 | 98.94 | 94.74 | 94.20 | IIP-Net54-C-GAP | 95.48 | 96.68 | 96.10 | 99.31 | 96.39 | 95.90 | IIP-Net105-GAP-FC | 99.55 | 99.31 | 99.31 | 99.65 | 99.31 | 99.50 | IIP-Net105-FC | 95.24 | 97.24 | 95.00 | 99.44 | 96.11 | 94.30 | IIP-Net105-C-GAP | 95.98 | 97.25 | 95.40 | 99.44 | 96.32 | 95.40 | IIP-Net156-GAP-FC | 99.55 | 99.28 | 99.28 | 99.64 | 99.28 | 99.50 | IIP-Net156-FC | 93.81 | 92.34 | 94.00 | 98.38 | 93.16 | 94.40 | IIP-Net156-C-GAP | 95.28 | 97.25 | 93.80 | 99.13 | 95.49 | 94.40 |
|
|